Biostatistics Network Seminars
Usual time: Wednesdays 16:00 - 17:00
Location: Room 102, Department of Statistical Science, 1-19 Torrington Place (1st floor).
Some seminars are held in different locations at different times. Click on the abstract for more details.
The Analysis of Binary Outcome Measures in Trials with Partial Nesting
A partially nested trial design arises in trials of complex interventions where patients are clustered in one treatment arm but not in the comparator. Examples include trials comparing activity classes or therapeutic support groups or therapist treatments with a placebo treatment or pharmacological control. After reviewing methods for continuous outcome measures, the analysis for binary outcome data will be considered. For binary data possible methods of analysis considered include an adjusted test of proportions, summary measures analyses, logistic models with either random effects, and logistic GEE models with an changeable correlation structure. The presentation will describe the challenges and results of simulations studies comparing the asymptotic and small sample properties of these method. Some more general issues regarding the trials with clustering will also be discussed.
Title and abstract to follow.
Page last modified on 27 mar 14 13:12